Heart Rate Profile During Exercise as a Predictor of Sudden Death

abstract & commentary

By John P. DiMarco, MD, PhD

Synopsis: Data are consistent with the concept that autonomic imbalance predisposes persons to life-threatening arrhythmias.

Source: Jouven X, et.al. Heart-Rate Profile During Exercise as a Predictor of Sudden Death. N Engl J Med. 2005;352:1951-1958.

The Paris prospective study performed exercise testing in male Parisian civil servants between the ages of 42 and 53 years. Subjects were entered, and the exercise tests were performed between 1967 and 1972. This paper discusses the relationship between heart rate profiles during exercise to subsequent sudden cardiac death. In the Paris Prospective Study, subjects with known or suspected cardiovascular disease were excluded and did not undergo the stress test. Patients with resting systolic blood pressure of more than 180 or abnormal electrocardiogram were also excluded. A total of 6456 men completed exercise testing and formed the study cohort. For the purpose of this analysis, however, 271 patients who manifest an ischemic response to exercise and 117 subjects who had an impaired chronotropic response (defined as inability to achieve 80% of predicted maximum heart rate) were excluded from the analysis described in this report. The exercise protocol used was a bicycle exercise test with 3 successive workloads: 2 minutes at 82 watts, 6 minutes at 164 watts, and 2 minutes at 191 watts. Heart rate was recorded at rest, before exercise, every 2 minutes during exercise, at peak exercise, and every minute during recovery. The test could be terminated because of fatigue, dyspnea, leg discomfort, chest pain, a systolic blood pressure of more than 250 mm Hg, a heart rate of more than 180 bpm, ventricular tachycardia, or ischemic electrocardiographic changes. During follow-up, deaths were tabulated, and causes of death were obtained from death certificates. A total of 355 subjects were lost to follow-up (4.6% of the original study cohort). The chi-square test for trends was used for comparisons among quintiles of heart rate. The relative risk of death was estimated with a Cox proportional hazard’s model for each quintile of heart rate.

A total of 5713 men composed the study group. They were followed for a mean of 23 years. There were 1516 deaths, including 400 deaths from cardiac causes. Of the latter, 81 were sudden cardiac deaths and 129 were nonsudden deaths. Jouven and colleagues describe the sudden deaths as "sudden deaths from myocardial infarction," but it appears they included all sudden deaths in this category. Subjects who suffered sudden death were older, had a higher body mass index, used more tobacco, had a higher resting heart rate, a higher systolic blood pressure, a higher serum total triglyceride and total cholesterol, and a higher prevalence of diabetes than controls. They also had a shorter duration of exercise and were less physically active. It was noted that sudden death victims more frequently had a parental history of sudden death. The duration of exercise was 6.0 ± 2.3 minutes in the sudden death group, 6.7 ± 2.6 minutes in the nonsudden death group, and 7.3 ± 2.5 minutes in the control group.

The influence of several heart rate variables on prognosis was then examined. After adjustment for clinical factors, the risk of sudden death increased progressively with the resting heart rate. In comparison with the risk in the lowest resting heart rate quintile, the risk in the highest quintile was 3.5 times higher. For nonsudden death, the risk in the highest heart rate quintile was 1.5 times that in the lowest quintile. Subjects with a heart rate increase of less than 89 bpm during exercise had 4 times the risk of sudden death, 1.2 times the risk of nonsudden death, and 1.5 times the risk of death from any cause.

The rate of heart rate recovery was also significantly related to the risk of sudden death. Compared to those with a heart rate decrease from maximum heart rate at 1 minute of more than 40 bpm (the highest quintile), subjects with a heart rate recovery of less than 25 bpm (the lowest quintile) had 2.1 times the risk of sudden death, and 0.9 times the risk of nonsudden cardiac death. In the Cox proportional hazard’s model, the heart rate profile was strongly associated with sudden death and less strongly associated with death from any cause, but showed no association with nonsudden death from myocardial infarction. The strongest predictor of sudden death was a lower increase in heart rate. This was true even when the analysis was restricted to subjects who stopped exercise at the 164-W level.

Jouven et al conclude that their data are consistent with the concept that autonomic imbalance predisposes persons to life-threatening arrhythmias. They believe that their data show that a greater risk of sudden death is associated with an impaired ability to increase not only vagal but also sympathetic activity to appropriate levels. This could be explained by reduced baro-reflex sensitivity. They suggest that one of the beneficial effects of regular exercise training may be improvement in baroreflex sensitivity and in autonomic milieux, which would favor survival in the setting of myocardial insult.


These data are from a very long-term trial performed in Parisian civil servants. Jouven et al have previously reported that detection of both ischemia and ventricular premature depolarizations during exercise are each independent predictors of subsequent sudden death during follow-up. Patients with ischemia were excluded in this study, but it is not reported if the role of ventricular ectopy was analyzed here. The current data do provide some potentially useful insights into the mechanisms of sudden death. It appears that abnormalities in both vagal and the sympathetic nervous system may be involved in increasing someone risk for future sudden death. Unfortunately, however, the use of these data in an individual is quite limited. Despite the long duration of this trial, the sudden death rate per year was still quite low. However, the more general observation that increased physical activity should improve sudden death rates can still be made.